NAFeb 4, 2019
Higher order multipoint flux mixed finite element methods on quadrilaterals and hexahedraIlona Ambartsumyan, Eldar Khattatov, Jeonghun Lee et al.
We develop higher order multipoint flux mixed finite element (MFMFE) methods for solving elliptic problems on quadrilateral and hexahedral grids that reduce to cell-based pressure systems. The methods are based on a new family of mixed finite elements, which are enhanced Raviart-Thomas spaces with bubbles that are curls of specially chosen polynomials. The velocity degrees of freedom of the new spaces can be associated with the points of tensor-product Gauss-Lobatto quadrature rules, which allows for local velocity elimination and leads to a symmetric and positive definite cell-based system for the pressures. We prove optimal $k$-th order convergence for the velocity and pressure in their natural norms, as well as $(k+1)$-st order superconvergence for the pressure at the Gauss points. Moreover, local postprocessing gives a pressure that is superconvergent of order $(k+1)$ in the full $L^2$-norm. Numerical results illustrating the validity of our theoretical results are included.
85.5NAMar 22
Error Analysis of the Explicit Splitting Scheme for Fluid-Poroelastic Structure Interaction ProblemsYifan Wang, Jeonghun Lee, Suncica Canic
We present a priori error analysis for a fully discrete, parallelizable, explicit loosely coupled scheme for the time-dependent Stokes-Biot problem. The method decouples the fluid and poroelastic subproblems in a fully explicit fashion, allowing each problem to be solved independently at each time step, with a consistent treatment of the interface conditions that provides stability and convergence of the scheme. The error analysis is carried out in a discrete energy framework. More specifically, we introduce Ritz-type projections in each subdomain, and subtract the fully discrete scheme from the time-discrete continuous formulation. This yields reduced error equations in which the dominant interpolation contributions cancel. The remaining consistency terms stem primarily from time discretization residuals and lagged interface data inherent to the explicit splitting. The main result of this manuscript is the derivation of a discrete error energy identity, and establishment of unconditional error estimates in a combined energy-dissipation norm via a Gronwall-type argument. These estimates demonstrate first-order accuracy in time and optimal spatial convergence rates, as determined by the degree of the finite element polynomials. Numerical experiments based on a manufactured solution corroborate the theory, confirming first-order temporal convergence for all variables, and spatial convergence orders consistent with the chosen approximation spaces.